OpenAI and Microsoft are facing a new lawsuit from the Center for Investigative Reporting (CIR). The CIR says that OpenAI has used its published content from Mother Jones and Reveal websites to train earlier versions of ChatGPT.
Also read: OpenAI scraps ChatGPT voice after Scarlett Johansson controversy
The CIR, the plaintiff says that OpenAI used its content without permission or a promise of compensation. CIR, founded in 1977, operates the nonprofit American magazine Mother Jones and Reveal podcast.
CIR sues OpenAI and Microsoft over copyright infringement
The Center for Investigative Reporting (CIR) filed the lawsuit in federal court in New York on Thursday. The nonprofit organization accused OpenAI and Microsoft of using their content without permission or compensation. The CIR says OpenAI has violated copyright laws by using its content to train ChatGPT.
The CEO of CIR Monika Bauerlein said, “This free rider behavior is not only unfair, it is a violation of copyright. The work of journalists, at CIR and everywhere, is valuable, and OpenAI and Microsoft know it.”
🚨[AI copyright lawsuit] The Center for Investigative Reporting (behind @MotherJones & @reveal) sues OpenAI & Microsoft for copyright infringement. Quotes:
"Defendants copied, used, abridged, and displayed CIR’s valuable content without CIR’s permission or authorization, and… pic.twitter.com/SeZ0VtDOMY
— Luiza Jarovsky (@LuizaJarovsky) June 29, 2024
In the official complaint, the plaintiff hired a data scientist to analyze the OpenWebText database. OpenWebText is an approximation of WebText, which is a corpus of scraped web pages created by OpenAI. The data scientist found that the dataset contains 17,434 URLs from Mother Jones and 415 from Reveal. OpenWebText and WebText have slightly different numbers of Mother Jones articles because the scraping process happened on different days.
The plaintiff said in the official complaint,
“When they populated their training sets with works of journalism, Defendants had a choice: to respect works of journalism, or not. Defendants chose the latter”
Also read: OpenAI’s chief scientist, Ilya Sutskever, bids farewell
OpenAI used two algorithms, Dragnet and Newspaper, to build the WebText database. Dragnet is designed to separate the main article content from other parts of the website, such as the header, footer, title, author name, and copyright notices. When OpenAI scrapped Mother Jones’ website, it removed anything in its footer and header. Additionally, the ChatGPT maker removed the copyright notice and terms of use information as per the complaint.
Furthermore, the plaintiff claims Microsoft knew that the scraped URLs had journalism content without author names, titles, and copyright notices, facilitating copyright infringement by Bing AI and ChatGPT.
The CIR is seeking profits from OpenAI and Microsoft and actual or statutory damages. The amount stated is a minimum of $750 per infringed work and $2,500 per DMCA violation.
OpenAI faces lawsuits from other publications
This is not the first lawsuit filed against OpenAI for copyright infringement. Since the release of ChatGPT in late 2022, OpenAI and Microsoft have faced numerous lawsuits from big names like the New York Times, The Intercept, the New York Daily News, and the Chicago Tribune.
Also read: Employees claim OpenAI and Google DeepMind hiding AI risks
Meanwhile, some major publishers and internet giants have signed licensing agreements with OpenAI, giving it access to their archives. These include TIME Magazine, News Corp, the Financial Times, Vox Media, the Associated Press, The Atlantic, Stack Overflow, and Reddit.
Cryptopolitan reporting by Randa Moses
